- Email: [email protected]
An epidermal growth factor-like toxin and two sodium channel toxins from the sea anemone Stichodactyla gigantea
Toxicon 41 (2003) 229–236 www.elsevier.com/locate/toxicon
An epidermal growth factor-like toxin and two sodium channel toxins from the sea anemone Stichodactyla gigantea Kazuo Shiomia,*, Tomohiro Honmaa, Masao Idea, Yuji Nagashimaa, Masami Ishidaa, Makoto Chinob a
Department of Food Science and Technology, Tokyo University of Fisheries, Konan-4, Minato-ku, Tokyo 108-8477, Japan b Pharmaceutical Group, Nippon Kayaku, Co., Ltd, Shimo-3, Kita-ku, Tokyo 115-8588, Japan Received 16 July 2002; accepted 11 September 2002
Abstract Three peptide toxins (gigantoxins I– III) with crab toxicity were isolated from the sea anemone Stichodactyla gigantea by gel filtration on Sephadex G-50 and reverse-phase HPLC on TSKgel ODS-120T and their complete amino acid sequences were determined. Gigantoxins II (44 residues) and III (48 residues) have LD50 (against crabs) of 70 and 120 mg/kg, respectively, and are analogous to the known type 1 and 2 sea anemone sodium channel toxins, respectively. On the other hand, gigantoxin I (48 residues) is potently paralytic to crabs (ED50 215 mg/kg), although its lethality is very weak (LD50 . 1000 mg/kg). Interestingly, gigantoxin I has 31 – 33% homologies with mammalian epidermal growth factors (EGFs), with the same location of six cysteine residues. In accordance with the sequence similarity, gigantoxin I exhibits EGF activity as evidenced by rounding of A431 cells and tyrosine phosphorylation of the EGF receptor in the cells, although much less potently than human EGF. Gigantoxin I is the first example of EGF-like toxins of natural origin. q 2002 Elsevier Science Ltd. All rights reserved. Keywords: Amino acid sequence; Epidermal growth factor-like toxin; Gigantoxin; Sea anemone; Stichodactyla gigantea
1. Introduction Sea anemones contain the following three classes of proteinous or peptidic toxins: 15 – 21 kDa hemolysins (Kem, 1988), 3 – 5 kDa sodium channel toxins (Kem, 1988; Kem et al., 1990, 1996; Norton, 1991) and 3.5 – 6.5 kDa potassium channel toxins (Castan˜eda et al., 1995; Schweitz et al., 1995; Cotton et al., 1997; Gendeh et al., 1997; Minagawa et al., 1998; Diochot et al., 1998). The most thoroughly studied toxins are sodium channel toxins that bind to the receptor site 3 of sodium channels in the excitable membrane, thereby prolonging the open state of the channels during the depolarization procedure. Because of this unique biological effect, several sea anemone sodium channel toxins have been on the market and utilized as * Corresponding author. Tel.: þ81-3-5463-0601; fax: þ 81-35463-0669. E-mail address: [email protected] (K. Shiomi).
valuable pharmacological reagents in many laboratories of the world. Sea anemone sodium channel toxins are classified into three types (types 1 –3), based on their primary structure (Norton, 1991; Kem et al., 1996). Regardless of the type, many of the known sea anemone sodium channel toxins are more lethal to crustaceans than to mammals. Using crab toxicity as an index, we have already isolated nearly 10 peptide toxins from several species of sea anemones (Shiomi et al., 1995, 1996, 1997; Lin et al., 1996; Ishida et al., 1997). Of the isolated toxins, two toxins (AETX II and III) from Actinia erythraea are quite distinct from the known sea anemone toxins (Shiomi et al., 1997). They are composed of 59 residues including 10 cysteine residues, having sequence similarities with neurotoxins from the Brazilian ‘armed’ spider Phoneutria nigriventer. Structurally novel peptide toxins, such as AETX II and III, are expected to still exist in sea anemones, which have not been studied. Indeed, we have recently found an unusual toxin
0041-0101/03/$ - see front matter q 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 0 4 1 - 0 1 0 1 ( 0 2 ) 0 0 2 8 1 - 7
230
K. Shiomi et al. / Toxicon 41 (2003) 229–236
(gigantoxin I), together with other two toxins (gigantoxins II and III) analogous to the known sea anemone sodium channel toxins, in Stichodactyla gigantea, a relatively large species found in tropical or subtropical waters. Very interestingly, gigantoxin I was weakly lethal but potently paralytic to crabs and had sequence similarities to epidermal growth factors (EGFs) from mammals. We report here the isolation, primary structures and biological activities of three peptide toxins (gigantoxins I– III) from S. gigantea.
2. Materials and methods 2.1. Sea anemone Specimens of S. gigantea were collected along the coasts of Kuroshima Island, Okinawa Prefecture, in October 1989. They were shipped frozen to our laboratory and stored at 2 20 8C until used. 2.2. Isolation method Frozen samples (five specimens, 285 g) were thawed and well macerated in a motor. A small aliquot (usually 5 g) of the macerate was homogenized in five volumes of distilled water and centrifuged at 18800 £ g for 15 min. The supernatant obtained was applied to gel filtration on a Sephadex G-50 column (2.5 £ 90 cm; Amersham Pharmacia Biotech, Buckinghamshire, UK), which was eluted with 0.15 M NaCl in 0.01 M phosphate buffer (pH 7.0). Fractions of 8 ml were collected and measured for absorbance at 280 nm and crab toxicity. Toxic fractions were pooled and subjected to reverse-phase HPLC on a TSKgel ODS-120T column (0.46 £ 25 cm; Tosoh, Tokyo, Japan). The column was washed with 0.1% trifluoroacetic acid (TFA) and then eluted with two steps of linear gradients of acetonitrile (0 – 17.5% in 5 min and 17.5 – 42% in 60 min) in 0.1% TFA at a flow rate of 1 ml/min. Peptides were monitored at 220 nm with a UV detector. The eluate corresponding to each toxin peak was manually collected, lyophilized and dissolved in desired solvents for subsequent experiments. 2.3. Assay of toxicity Lethal or paralytic activity was assayed using freshwater crabs (Potamon dehaani ) weighing about 5 g purchased from the Tokyo Central Wholesale Market and male mice (ddY strain) weighing about 20 g from Sankyo Labo Service (Tokyo, Japan). Sample solutions were injected into crabs at the junction between the body and the leg or injected i.v. into mice. The injection volume was fixed at 10 ml/g of body weight of crabs or mice. In order to calculate LD50 (for lethal activity) or ED50 (for paralytic activity) against crabs by the method of Litchfield and Wilcoxon (1949), groups of five animals were challenged with various doses of toxin and observed for mortality up to 2 h. For comparison,
human EGF (Sigma, St Louis, USA) was examined for toxicity against crabs and mice. 2.4. Assay of EGF activity EGF activity was assessed using human epidermoid carcinoma A431 cells, from the viewpoints of morphological change and tyrosine phosphorylation of the EGF receptor (EGFR). A431 cells were cultured in a 5% CO2 atmosphere in Dulbecco’s modified Eagle’s medium (DMEM), supplemented with calf serum (10%), kanamycin (0.06 mg/ml), streptomycin (0.1 mg/ml) and penicillin (10 units/ml). For morphological observations, A431 cells were plated at 1 £ 105 in 1 ml of DMEM in 6-well plates (f 35 mm) and allowed to attach at 37 8C for 18 –20 h. Subsequently, the cells were washed twice with 1 ml of PBS (composed of 137 mM NaCl, 2.7 mM KCl, 8 mM Na2HPO4 and 1.5 mM KH2PO4) and incubated in 1 ml of PBS, with or without test peptide (10, 100 or 1000 nM gigantoxin I or 0.01, 0.1, 1 or 10 nM human EGF), at 37 8C for 30 min. Morphological changes of the cells were examined by phase-contrast microscopy. For the assay of tyrosine phosphorylation of the EGFR, A431 cells (1 £ 106 cells) were grown in 1 ml of DMEM in 10 cm dishes at 37 8C for 18– 20 h. Following incubation with test peptide at 37 8C for 30 min, the cells were washed twice with cold PBS (supplemented with 0.9 mM CaCl2 and 0.5 mM MgCl2), scraped with a cell scraper into 1 ml of the same cold PBS and centrifuged at 1000 £ g and 4 8C for 4 min. The cell pellet, after being stored at 2 80 8C for 18 – 20 h, was homogenized in 1 ml of radioimmune precipitation (RIPA) buffer (25 mM HEPES buffer, pH 7.8, containing 1.5% Triton-X100, 0.1% SDS, 500 mM NaCl, 100 mM Na3VO4, 50 mM NaF, 5 mM EDTA, 1 mM phenylmethanesulfonyl fluoride, 0.01% leupeptin and 1% sodium deoxycholate) at 4 8C for 30 min and centrifuged at 15000 £ g and 4 8C for 15 min. The supernatant was mixed with 25 ml of 50% Protein G Sepharose 4B (Amersham Pharmacia Biotech) suspension in RIPA buffer and kept with stirring at 4 8C for 1 h. After centrifugation at 2000 £ g and 4 8C for 4 min, the supernatant was added with 2 mg of mouse anti-EGFR antibody (Santa Cruze Biotechnology, Santa Cruze, USA) and 25 ml of 50% Protein G Sepharose 4B suspension in RIPA buffer. The mixture was stirred at 4 8C for 3 h and centrifuged at 2000 £ g and 4 8C for 4 min. After being washed four times with 1 ml of RIPA buffer, the precipitate was suspended in 25 ml of electrophoresis loading buffer (0.52 M Tris – HCl buffer, pH 6.8, containing 2.15% SDS, 5% 2-mercaptoethanol, 7% glycerol and 0.0025% bromophenol blue; Toyobo, Tokyo, Japan), heated at 95 8C for 3 min and centrifuged briefly. A 20 ml portion of the supernatant was subjected to SDS – PAGE on a 5% polyacrylamide gel, which was performed by the method of Laemmli (1970). Proteins separated by SDS – PAGE were transferred onto a nitrocellulose membrane at 100 V for 1 h.
K. Shiomi et al. / Toxicon 41 (2003) 229–236
The membrane was blocked with 10% skimmed milk in TBS-Tween (10 mM Tris – HCl buffer, pH 7.5, containing 0.1% Tween 20) at room temperature for 1 h and washed with TBS-Tween. The transfer was probed with mouse antiphosphotyrosine antibody 4G10 (1:5000 dilution; Seikagaku Corp., Tokyo, Japan) at room temperature for 2 h, followed by anti-mouse IgG antibody RPN 1001 conjugated with horseradish peroxidase (1:1000 dilution; Amersham Pharmacia Biotech) at room temperature for 1 h. After rinsing with TBS, labeled proteins were visualized with an ECL detection system (Amersham Pharmacia Biotech). 2.5. Peptide determination Peptides were determined by the method of Lowry et al. (1951) using bovine serum albumin as a standard. 2.6. Reduction and S-pyridylethylation
231
3. Results 3.1. Isolation of gigantoxins In gel filtration on Sephadex G-50, toxins with crab lethality appeared between fractions 30 and 42 (Fig. 1(A)). When the toxic fraction obtained by gel filtration was subjected to reverse-phase HPLC on TSKgel ODS-120T, three toxins (named gigantoxins I, II and III in the order of elution) were eluted at retention times of 31, 39 and 50 min, respectively (Fig. 1(B)). In a typical run, 5 g of the starting sample yielded as much as 5960 mg of gigantoxin I, 150 mg of gigantoxin II and 250 mg of gigantoxin III. 3.2. Amino acid sequences of gigantoxins The entire amino acid sequences of gigantoxins I (48 residues), II (44 residues) and III (48 residues) were determined based on the strategies shown in Fig. 2.
Each purified gigantoxin (200 – 300 mg) was dissolved in 300 ml of 0.5 M Tris – HCL buffer (pH 8.5) containing 7 M guanidine hydrochloride and 10 mM EDTA and reduced with 1 mg of dithiothreitol at room temperature for 2 h. The reaction mixture was then added with 2 ml of 4-vinylpyridine and kept in the dark at room temperature for 1.5 h. The pyridylethylated (PE)-toxin was purified by reverse-phase HPLC on TSKgel ODS-120T with a linear gradient of acetonitrile (0 –70% in 40 min) in 0.1% TFA. 2.7. Enzyme digestion PE-gigantoxins (100 mg each) were individually digested with 40 mg of carboxypeptidase W (EC 3.4.16.1; Seikagaku Corp.) in 120 ml of 0.05 M acetate buffer (pH 4.0) at 30 8C. Aliquots of the reaction mixture were withdrawn at suitable intervals and subjected to an MLC-703 amino acid analyzer (Atto, Tokyo, Japan). PE-gigantoxin II (70 mg) was also digested with 5 mg of chymotrypsin (EC 3.4.21.1; Roche, Mannheim, Germany) in 500 ml of 0.1 M Tris – HCl buffer (pH 7.8) containing 0.01 M CaCl2 at 25 8C for 18 h and PE-gigantoxin III (50 mg) with 10 mg of V8 protease (EC 3.4.21.19; ICN Biomedicals, Irvine, USA) in 250 ml of 0.1 M ammonium hydrogencarbonate containing 2 mM EDTA at 37 8C for 18 h. Peptide fragments produced were isolated for sequencing by reverse-phase HPLC on TSKgel ODS-120T with a linear gradient of acetonitrile (0– 50% in 80 min) in 0.1% TFA. 2.8. Sequence analysis Amino acid sequence analyses were performed with an automatic gas-phase protein sequencer (LF-3400D TriCart with high sensitivity chemistry; Beckman Coulter, Fullerton, USA).
Fig. 1. Isolation of gigantoxins. (A) Gel filtration on Sephadex G50. Toxic fractions are indicated by a bar. (B) Reverse-phase HPLC on TSKgel ODS-120T. Gigantoxins I– III were eluted in labeled peaks.
232
K. Shiomi et al. / Toxicon 41 (2003) 229–236
Fig. 2. Strategies for the determination of entire amino acid sequences of gigantoxins I– III. Labels above the straight lines with arrows at both ends: Native, N-terminal sequences determined by sequencing of the native peptides; CpW, C-terminal sequences determined by the carboxypeptidase W digestion method; Chy, sequences of peptide fragments from the chymotrypsin digest of PE-gigantoxin II; V8, sequence of a peptide fragment from the V8 protease digest of PE-gigantoxin III.
Sequencing of the native peptides identified N-terminal 46, 35 and 45 residues for gigantoxins I, II and III, respectively. The C-terminal sequences were deduced to be 44EQMSV48 for gigantoxin I, 43KQ44 for gigantoxin II and 46RPR48 for
gigantoxin III by the estimation of amino acids released upon carboxypeptidase W digestion of the PE-derivatives. In the case of gigantoxin I, the N- and C-terminal sequences partly overlapped with each other. The remaining sequence
Fig. 3. Alignment of the amino acid sequence of gigantoxin I with those of mammalian EGFs (A), gigantoxin II with type 1 sea anemone sodium channel toxins (B) and gigantoxin III with type 2 sea anemone sodium channel toxins (C). (A) The residues identical with gigantoxin I are boxed. (B) AFT-I is from Anthopleura fuscoviridis (Sunahara et al., 1987); AP-A from A. xanthogrammica (Tanaka et al., 1977); ATX I from A. sulcata (Wunderer and Eulitz, 1978); Cp I from C. passiflora (Shiomi et al., 1995); and Rc I from Radianthus (Heteractis ) crispus (Shiomi et al., 1996). The residues identical with gigantoxin II are boxed. Asterisks under the sequences of Rc I represent the residues typical of type 1 toxins. Hydroxyproline at position 3 of Cp I and Rc I is indicated by ‘O’. (C) Rp II from R. paumotensis (Wemmer et al., 1986); RTX I (Zykova and Kozlovskaya, 1989), II (Zykova et al., 1988a) and V (Zykova et al., 1988b) from R. macrodactylus; and Sh I from S. helianthus (Kem et al., 1989). Asterisks under the sequences of Sh I represent the residues typical of type 2 toxins.
K. Shiomi et al. / Toxicon 41 (2003) 229–236
of gigantoxin II was compensated by two peptide fragments (corresponding to the segments 32 –41 and 42 – 44) isolated from the chymotrypsin digest of the PE-derivative. Although not confirmed, these two fragments were judged to be linked together, since, as described below, gigantoxin II is a member of the type 1 sea anemone sodium channel toxins with two successive cysteine residues in the Cterminal region. As for gigantoxin III, one of the peptide fragments isolated from the V8 protease digest of the PEderivative was unequivocally the C-terminal peptide (corresponding to the segment 29 –48) overlapping with the N-terminal sequence. Very surprisingly, a search by protein sequence databases revealed that gigantoxin I is very similar to mammalian EGFs. For comparison, the amino acid sequence of gigantoxin I is aligned with those of EGFs from human (Gregory and Preston, 1977), pig (Pascall et al., 1991) and mouse (Savage et al., 1972) in Fig. 3(A). As much as 31 – 33% homologies are recognized between gigantoxin I and EGFs; the location of six cysteine residues in gigantoxin I is completely identical with that in EGFs. On the other hand, gigantoxins II and III are obviously analogous to type 1 and 2 toxins, respectively, when aligned with the following known sea anemone sodium channel toxins (Fig. 3(B) and (C)): type 1 toxins, AFT-I from Anthopleura fuscoviridis (Sunahara et al., 1987), AP-A from Anthopleura xanthogrammica (Tanaka et al., 1977), ATX I from Anemonia sulcata (Wunderer and Eulitz, 1978), Cp I from Condylactis passiflora (Shiomi et al., 1995) and Rc I from Radianthus (Heteractis ) crispus (Shiomi et al., 1996); type 2 toxins, Rp II from Radianthus paumotensis (Wemmer et al., 1986), RTX I (Zykova and Kozlovskaya, 1989), II (Zykova et al., 1988a) and V (Zykova et al., 1988b) from Radianthus macrodactylus and Sh I from Stichodactyla helianthus (Kem et al., 1989). The residues typical of type 1 and two toxins are well conserved in gigantoxins II and III, respectively. It is, however, noticeable that gigantoxin II has a rather shorter chain than the known type 1 toxins and that gigantoxin III has an unusual C-terminal sequence (46RPR48) as compared to the known type 2 toxins.
233
human EGF, any ill signs including paralysis were not observed in both mice and crabs even at 1000 mg/kg. EGF, in Ca2þ free media, is known to change the polygonal pavement-like shape (Fig. 4(A)) of A431 cells to the round shape usually within 10 min (Chinkers et al., 1981). In this study, therefore, rounding of the cells was judged 30 min after treatment with human EGF or gigantoxin I. As a result, human EGF caused rounding of almost all A431 cells at 1 and 10 nM (data not shown), while gigantoxin I showed the same effect at a 100– 1000 times higher concentration (1000 nM) (Fig. 4(B)). No morphological changes were observed at less than 0.1 nM human EGF and less than 100 nM gigantoxin I. Furthermore, gigantoxin I was found to induce tyrosine phosphorylation of the EGFR in A431 cells. When the EGFR was purified from the cells incubated with the toxin by use of the anti-EGFR antibody and subjected to Western blotting, the 170 kDa protein corresponding to the EGFR reacted with the anti-phosphotyrosine antibody (Fig. 5). The intensity of the visualized band at 500 or 1000 nM gigantoxin I was almost the same as that at 1 nM human EGF, indicating that gigantoxin I is 500– 1000 times less in the potency of inducing autophosphorylation of the EGFR than human EGF.
3.3. Biological activities of gigantoxins None of the three gigantoxins caused any symptoms in mice at a dose of 1000 mg/kg. As for crab toxicity, both gigantoxins II and III were potently lethal with LD50 of 70 and 120 mg/kg, respectively. In contrast, gigantoxin I induced tonic paralysis in crabs with an ED50 of 215 mg/kg, while it killed only one of the five crabs challenged with a dose of 1000 mg/kg, indicating that its LD50 is over 1000 mg/kg. The crabs injected with effective doses near ED50 exhibited paralysis after 10 – 20 min, which made them impossible to walk for 5 – 10 min. At 1000 mg/kg, paralysis appeared immediately after injection and lasted for a longer time (30 –60 min). In the case of
Fig. 4. Morphological change of A431 cells by gigantoxin I. The cells were incubated at 37 8C for 30 min in PBS without (A) or with gigantoxin I (1000 nM) (B).
234
K. Shiomi et al. / Toxicon 41 (2003) 229–236
Fig. 5. Tyrosine phosphorylation of EGFR in A431 cells by human EGF and gigantoxin I. After the cells were incubated with human EGF or gigantoxin I at 37 8C for 30 min, the EGFR was purified from the cells and subjected to Western blot analysis. Lanes: 1, 10 nM human EGF; 2, 1 nM human EGF; 3, 1000 nM gigantoxin I; 4, 500 nM gigantoxin I; 5, 100 nM gigantoxin I; 6, control.
4. Discussion In this study, three peptide toxins (gigantoxins I– III) with crab toxicity were isolated from the sea anemone S. gigantea and their complete amino acid sequences were determined. One peptide toxin with crab lethality has previously been isolated from the same sea anemone (formerly called Stoichactis giganteus ) but its amino acid sequence has not been elucidated (Schweitz et al., 1981). Judging from the reported amino acid composition, this peptide toxin seems to be the same as gigantoxin III. Of the three peptide toxins from S. gigantea, gigantoxin I is of particular interest in that it has high homologies (31 – 33%) with mammalian EGFs with the same location of six cysteine residues. In accordance with the sequence similarity, gigantoxin I induces the rounding of A431 cells and tyrosine phosphorylation of the EGFR in the cells, although much less potently than human EGF. However, gigantoxin I is distinguishable from EGFs in crab toxicity; gigantoxin I is weakly lethal with LD50 . 1000 mg/kg and causes paralysis with an ED50 of 215 mg/kg, while human EGF is non-toxic even at 1000 mg/kg. As far as we know, gigantoxin I is the first example of EGF-like toxins of natural origin. EGF-like toxins such as gigantoxin I may be contained in sea anemones other than S. gigantea. In future survey of EGF-like toxins in sea anemones, however, it should be kept in mind that they may be paralytic but not potently lethal to crabs, as observed with gigantoxin I. According to Carpenter and Cohen (1990), EGFs and EGF-like molecules such as transforming growth factor-a have the same residues at 11 positions, i.e. Cys-6 (numbering is based on the sequence of gigantoxin I shown in Fig. 2), Cys-15, Gly-19, Cys-21, Cys-32, Cys-34, Tyr-38, Gly-40, Arg-42, Cys-43 and Leu-48. All of these residues, except for Leu-48, are also well conserved in gigantoxin I. It should be noted that Leu-48 has previously
been identified as one of the essential residues for the biological activity of EGFs (Engler et al., 1988; Ray et al., 1988). Thus, the absence of Leu-48 seems to explain the reason why gigantoxin I is much less potent in EGF activity (rounding of A431 cells and tyrosine phosphorylation of the EGFR) than human EGF. It is also interesting to note that gigantoxin I has marked alterations including the absence of Leu-48 in the C-terminal portion as compared to EGFs. This may be the molecular basis for the crab toxicity of gigantoxin I. Sea anemone toxins are generally considered to be contained in specialized stinging cells called nematocysts. This is probably the case with gigantoxins, although not confirmed. In relation to this, it is worth mentioning that the content of gigantoxin I in S. gigantea is assumed to be more than 0.1% on the basis of yield (5960 mg from 5 g of wet sample), being much higher than those of gigantoxins II and III. Although gigantoxin I is substantially non-lethal to crabs, its high content and potent paralytic activity against crabs suggest that it functions in S. gigantea as a more powerful weapon to paralyze prey animals than the other two peptide toxins. The finding of an EGF-like toxin (gigantoxin I) in S. gigantea is also of interest from the viewpoint of molecular evolution of EGFs. So far, there have been numerous studies on EGFs from wide species of animals, especially vertebrates. Nevertheless, no information is available concerning EGFs in animals, such as sponges, corals and sea anemones, which are the nearest to the phylogenetic root of the animal kingdom. The present study clearly demonstrates that gigantoxin I belongs to the EGF family with respect to structure and activity. However, the significantly high content of gigantoxin I in S. gigantea is extremely disproportionate to the known regulatory molecules such as mammalian EGFs. This suggests that gigantoxin I seems not to be important as a growth factor for the sea anemone. It is thus tentatively assumed that the ancestors of EGFs originally had functioned as toxins as in the case of gigantoxin I and that they had lost toxic properties during the evolution process in the animal kingdom. In order to ascertain this fascinating assumption, it is essential to elucidate the gene structure of gigantoxin I and compare it with those of the known EGFs. The gene expression analysis probed with the gigantoxin I cDNA sequence is also necessary to know if gigantoxin I is the sole EGF-like molecule in the sea anemone or another EGF is present. As for gigantoxins II and III, they are analogous to the known sea anemone sodium channel toxins. According to the classification of sea anemone sodium channel toxins by Norton (1991), gigantoxins II and III are identified as type 1 and 2 toxins, respectively. It has previously been proposed that sea anemones belonging to the genera Anthopleura and Anemonia contain type 1 toxins, while those to the genera Radianthus (or Heteractis ) and Stichodactyla contain type 2 toxins (Norton, 1991). Radianthus crispus not having a type 2 toxin but a type 1 toxin has been the only exception
K. Shiomi et al. / Toxicon 41 (2003) 229–236
(Shiomi et al., 1996). Therefore, our results are valuable in showing first the coexistence of both type 1 and 2 toxins in one species of sea anemone. The relationship between the taxonomical position of sea anemones and the type of their sodium channel toxins will require future detailed study. Finally, it should be again emphasized that gigantoxin I is the first example of EGF-like molecules with both toxic and EGF activities. Its mode of action on crabs, structure – activity relationships and gene structure remain to be elucidated. Also, the distribution of similar EGF-like toxins in sea anemones should be surveyed.
Acknowledgements The authors thank Miss S. Kawahata for technical assistance. This study was partly supported by a Grant-inAid for Scientific Research from the Ministry of Education, Science, Sports and Culture of Japan.
References Carpenter, G., Cohen, S., 1990. Epidermal growth factor. J. Biol. Chem. 265, 7709–7712. Castan˜eda, O., Sotolongo, V., Amor, A.M., Sto¨klin, R., Anderson, ˚ ., Wernstedt, C., Karlsson, E., A.J., Harvey, A.L., Engstro¨m, A 1995. Characterization of a potassium channel toxin from the Caribbean sea anemone Stichodactyla helianthus. Toxicon 33, 603–613. Chinkers, M., Mckanna, J.A., Cohen, S., 1981. Rapid rounding of epidermoid carcinoma cells A-431 induced by epidermal growth factor. J. Cell Biol. 88, 422–429. Cotton, J., Crest, M., Bouet, F., Alessandri, N., Gola, M., Forest, E., Karlsson, E., Castan˜eda, O., Harvey, A.L., Vita, C., Me´nez, A., 1997. A potassium-channel toxin from the sea anemone Bunodosoma granulifera, an inhibitor for Kv1 channels. Revision of the amino acid sequence, disulfide-bridge assignment, chemical synthesis, and biological activity. Eur. J. Biochem. 244, 192–202. Diochot, S., Schweitz, H., Be´ress, L., Lazdunski, M., 1998. Sea anemone peptides with a specific blocking activity against the fast inactivating potassium channel Kv3.4. J. Biol. Chem. 273, 6744–6749. Engler, D.A., Matsunami, R.K., Campion, S.R., Stringer, C.D., Stevens, A., Niyogi, S.K., 1988. Cloning of authentic human epidermal growth factor as a bacterial secretary protein and its initial structure–function analysis by site-directed mutagenesis. J. Biol. Chem. 263, 12384–12390. Gendeh, G.S., Young, L.C., de Medeiros, L.C., Jeyaseelan, K., Harvey, A.L., Chung, M.C.M., 1997. A new potassium channel toxin from the sea anemone Heteractis magnifica: isolation, cDNA cloning, and functional expression. Biochemistry 36, 11461– 11471. Gregory, H., Preston, B.M., 1977. The primary structure of human urogastrone. Int. J. Pept. Protein Res. 9, 107–118. Ishida, M., Yokoyama, A., Shimakura, K., Nagashima, Y., Shiomi, K., 1997. Halcurin, a polypeptide toxin in the sea anemone
235
Halcurias sp with a structural resemblance to type 1 and 2 toxin. Toxicon 35, 537–544. Kem, W.R., 1988. Sea anemone toxins: structure and action. In: Hessinger, D.A., Lenhoff, H.M. (Eds.), The Biology of Nematocysts, Academic Press, New York, pp. 375 –405. Kem, W.R., Parten, B., Pennington, M.W., Price, D.A., Dunn, B.M., 1989. Isolation, characterization, and amino acid sequence of a polypeptide neurotoxin occurring in the sea anemone Stichodactyla helianthus. Biochemistry 28, 3483–3489. Kem, W.R., Pennington, M.W., Dunn, B.M., 1990. Sea anemone polypeptide toxins affecting sodium channels. In: Hall, S., Stricharz, G. (Eds.), Marine Toxins. Origin, Structure and Molecular Pharmacology, ACS, Washington, DC, pp. 279 –289. Kem, W.R., Pennington, M.W., Krafte, D.S., Hill, R.J., 1996. Sea anemone toxins affecting sodium channels: are the similarities than the differences? . In: Lazarovici, P., Spira, M.E., Zlotkin, E. (Eds.), Biochemical Aspects of Marine Pharmacology, Alakasn, Fort Collins, pp. 98–120. Laemmli, U.K., 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680 –685. Lin, X.Y., Ishida, M., Nagashima, Y., Shiomi, K., 1996. A polypeptide toxin in the sea anemone Actinia equina homologous with other sea anemone sodium channel toxins: isolation and amino acid sequence. Toxicon 34, 57–65. Litchfield, J.T. Jr, Wilcoxon, F., 1949. A simplified method of evaluating dose-effect experiments. J. Pharmacol. Exp. Ther. 96, 99–113. Lowry, O.H., Rosebrough, A.L., Farr, A.L., Randall, R.J., 1951. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193, 265–275. Minagawa, S., Ishida, M., Nagashima, Y., Shiomi, K., 1998. Primary structure of a potassium channel toxin from the sea anemone Actinia equina. FEBS Lett. 427, 149– 151. Norton, R.S., 1991. Structure and structure–function relationships of sea anemone proteins that interact with the sodium channel. Toxicon 29, 1051–1084. Pascall, J.C., Jones, D.S.C., Doel, S.M., Clements, J.M., Hunter, M., Fallon, T., Edwards, M., Brown, K.D., 1991. Cloning and characterization of a gene encoding pig epidermal growth factor. J. Mol. Endocrinol. 6, 63–70. Ray, P., Mey, F.J., Montelione, G.T., Liu, J.-F., Narang, S.A., Scherage, H.A., Wu, R., 1988. Structure–function studies of murine epidermal growth factor: expression and site-directed mutagenesis of epidermal growth factor gene. Biochemistry 27, 7289–7295. Savage, C.R. Jr, Inagami, T., Cohen, S., 1972. The primary structure of epidermal growth factor. J. Biol. Chem. 247, 7612–7621. Schweitz, H., Vincent, J.P., Barhanin, J., Frelin, C., Linden, G., Hugues, M., Lazdunski, M., 1981. Purification and pharmacological properties of eight sea anemone toxins from Anemonia sulcata, Anthopleura xanthogrammica, Stoichactis giganteus, and Actinodendron plumosum. Biochemistry 20, 5245–5252. Schweitz, H., Bruhn, T., Guillemare, E., Moinier, D., Lancelin, J.M., Be´ress, L., Lazdunski, M., 1995. Kalicludines and kaliseptine. Two different classes of sea anemone toxins for voltage-sensitive Kþ channels. J. Biol. Chem. 270, 25121–25126.
236
K. Shiomi et al. / Toxicon 41 (2003) 229–236
Shiomi, K., Lin, X.Y., Nagashima, Y., Ishida, M., 1995. Isolation and amino acid sequence of polypeptide toxins in the sea anemone Condylactis passiflora. Fish. Sci. 61, 1016–1021. Shiomi, K., Lin, X.Y., Nagashima, Y., Ishida, M., 1996. Isolation and amino acid sequence of a polypeptide toxin from the sea anemone Radianthus crispus. Fish. Sci. 62, 629– 633. Shiomi, K., Qian, W.H., Lin, X.Y., Shimakura, K., Nagashima, Y., Ishida, M., 1997. Novel polypeptide toxins with crab toxicity from the sea anemone Anemonia erythraea. Biochim. Biophys. Acta 1335, 191 –198. Sunahara, S., Muramoto, K., Tenma, K., Kamiya, H., 1987. Amino acid sequence of two sea anemone toxins from Anthopleura fuscoviridis. Toxicon 25, 211– 219. Tanaka, M., Haniu, M., Yasunobu, K.T., Norton, T.R., 1977. Amino acid sequence of the Anthopleura xanthogrammica heart stimulant anthopleurin-A. Biochemistry 16, 204–208.
Wemmer, D.E., Kumar, N.V., Metrione, R.M., Lazdunski, M., Drobny, G., Kallenbach, N.R., 1986. NMR analysis and sequence of toxin II from the sea anemone Radianthus paumotensis. Biochemistry 25, 6842– 6849. Wunderer, G., Eulitz, M., 1978. Amino acid sequence of toxin 1 from Anemonia sulcata. Eur. J. Biochem. 89, 11–17. Zykova, T.A., Kozlovskaya, E.P., 1989. Amino acid sequence of neurotoxin I from the sea anemone Radianthus macrodactylus. Bioorg. Khim. 15, 1301–1306. Zykova, T.A., Kozlovskaya, E.P., Elyakov, G.B., 1988a. Amino acid sequence of neurotoxin II from the sea anemone Radianthus macrodactylus. Bioorg. Khim. 14, 878 –882. Zykova, T.A., Kozlovskaya, E.P., Elyakov, G.B., 1988b. Amino acid sequence of neurotoxins IV and V from the sea anemone Radianthus macrodactylus. Bioorg. Khim. 14, 1489–1494.
An epidermal growth factor-like toxin and two sodium channel toxins from the sea anemone Stichodactyla gigantea Kazuo Shiomia,*, Tomohiro Honmaa, Masao Idea, Yuji Nagashimaa, Masami Ishidaa, Makoto Chinob a
Department of Food Science and Technology, Tokyo University of Fisheries, Konan-4, Minato-ku, Tokyo 108-8477, Japan b Pharmaceutical Group, Nippon Kayaku, Co., Ltd, Shimo-3, Kita-ku, Tokyo 115-8588, Japan Received 16 July 2002; accepted 11 September 2002
Abstract Three peptide toxins (gigantoxins I– III) with crab toxicity were isolated from the sea anemone Stichodactyla gigantea by gel filtration on Sephadex G-50 and reverse-phase HPLC on TSKgel ODS-120T and their complete amino acid sequences were determined. Gigantoxins II (44 residues) and III (48 residues) have LD50 (against crabs) of 70 and 120 mg/kg, respectively, and are analogous to the known type 1 and 2 sea anemone sodium channel toxins, respectively. On the other hand, gigantoxin I (48 residues) is potently paralytic to crabs (ED50 215 mg/kg), although its lethality is very weak (LD50 . 1000 mg/kg). Interestingly, gigantoxin I has 31 – 33% homologies with mammalian epidermal growth factors (EGFs), with the same location of six cysteine residues. In accordance with the sequence similarity, gigantoxin I exhibits EGF activity as evidenced by rounding of A431 cells and tyrosine phosphorylation of the EGF receptor in the cells, although much less potently than human EGF. Gigantoxin I is the first example of EGF-like toxins of natural origin. q 2002 Elsevier Science Ltd. All rights reserved. Keywords: Amino acid sequence; Epidermal growth factor-like toxin; Gigantoxin; Sea anemone; Stichodactyla gigantea
1. Introduction Sea anemones contain the following three classes of proteinous or peptidic toxins: 15 – 21 kDa hemolysins (Kem, 1988), 3 – 5 kDa sodium channel toxins (Kem, 1988; Kem et al., 1990, 1996; Norton, 1991) and 3.5 – 6.5 kDa potassium channel toxins (Castan˜eda et al., 1995; Schweitz et al., 1995; Cotton et al., 1997; Gendeh et al., 1997; Minagawa et al., 1998; Diochot et al., 1998). The most thoroughly studied toxins are sodium channel toxins that bind to the receptor site 3 of sodium channels in the excitable membrane, thereby prolonging the open state of the channels during the depolarization procedure. Because of this unique biological effect, several sea anemone sodium channel toxins have been on the market and utilized as * Corresponding author. Tel.: þ81-3-5463-0601; fax: þ 81-35463-0669. E-mail address: [email protected] (K. Shiomi).
valuable pharmacological reagents in many laboratories of the world. Sea anemone sodium channel toxins are classified into three types (types 1 –3), based on their primary structure (Norton, 1991; Kem et al., 1996). Regardless of the type, many of the known sea anemone sodium channel toxins are more lethal to crustaceans than to mammals. Using crab toxicity as an index, we have already isolated nearly 10 peptide toxins from several species of sea anemones (Shiomi et al., 1995, 1996, 1997; Lin et al., 1996; Ishida et al., 1997). Of the isolated toxins, two toxins (AETX II and III) from Actinia erythraea are quite distinct from the known sea anemone toxins (Shiomi et al., 1997). They are composed of 59 residues including 10 cysteine residues, having sequence similarities with neurotoxins from the Brazilian ‘armed’ spider Phoneutria nigriventer. Structurally novel peptide toxins, such as AETX II and III, are expected to still exist in sea anemones, which have not been studied. Indeed, we have recently found an unusual toxin
0041-0101/03/$ - see front matter q 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 0 4 1 - 0 1 0 1 ( 0 2 ) 0 0 2 8 1 - 7
230
K. Shiomi et al. / Toxicon 41 (2003) 229–236
(gigantoxin I), together with other two toxins (gigantoxins II and III) analogous to the known sea anemone sodium channel toxins, in Stichodactyla gigantea, a relatively large species found in tropical or subtropical waters. Very interestingly, gigantoxin I was weakly lethal but potently paralytic to crabs and had sequence similarities to epidermal growth factors (EGFs) from mammals. We report here the isolation, primary structures and biological activities of three peptide toxins (gigantoxins I– III) from S. gigantea.
2. Materials and methods 2.1. Sea anemone Specimens of S. gigantea were collected along the coasts of Kuroshima Island, Okinawa Prefecture, in October 1989. They were shipped frozen to our laboratory and stored at 2 20 8C until used. 2.2. Isolation method Frozen samples (five specimens, 285 g) were thawed and well macerated in a motor. A small aliquot (usually 5 g) of the macerate was homogenized in five volumes of distilled water and centrifuged at 18800 £ g for 15 min. The supernatant obtained was applied to gel filtration on a Sephadex G-50 column (2.5 £ 90 cm; Amersham Pharmacia Biotech, Buckinghamshire, UK), which was eluted with 0.15 M NaCl in 0.01 M phosphate buffer (pH 7.0). Fractions of 8 ml were collected and measured for absorbance at 280 nm and crab toxicity. Toxic fractions were pooled and subjected to reverse-phase HPLC on a TSKgel ODS-120T column (0.46 £ 25 cm; Tosoh, Tokyo, Japan). The column was washed with 0.1% trifluoroacetic acid (TFA) and then eluted with two steps of linear gradients of acetonitrile (0 – 17.5% in 5 min and 17.5 – 42% in 60 min) in 0.1% TFA at a flow rate of 1 ml/min. Peptides were monitored at 220 nm with a UV detector. The eluate corresponding to each toxin peak was manually collected, lyophilized and dissolved in desired solvents for subsequent experiments. 2.3. Assay of toxicity Lethal or paralytic activity was assayed using freshwater crabs (Potamon dehaani ) weighing about 5 g purchased from the Tokyo Central Wholesale Market and male mice (ddY strain) weighing about 20 g from Sankyo Labo Service (Tokyo, Japan). Sample solutions were injected into crabs at the junction between the body and the leg or injected i.v. into mice. The injection volume was fixed at 10 ml/g of body weight of crabs or mice. In order to calculate LD50 (for lethal activity) or ED50 (for paralytic activity) against crabs by the method of Litchfield and Wilcoxon (1949), groups of five animals were challenged with various doses of toxin and observed for mortality up to 2 h. For comparison,
human EGF (Sigma, St Louis, USA) was examined for toxicity against crabs and mice. 2.4. Assay of EGF activity EGF activity was assessed using human epidermoid carcinoma A431 cells, from the viewpoints of morphological change and tyrosine phosphorylation of the EGF receptor (EGFR). A431 cells were cultured in a 5% CO2 atmosphere in Dulbecco’s modified Eagle’s medium (DMEM), supplemented with calf serum (10%), kanamycin (0.06 mg/ml), streptomycin (0.1 mg/ml) and penicillin (10 units/ml). For morphological observations, A431 cells were plated at 1 £ 105 in 1 ml of DMEM in 6-well plates (f 35 mm) and allowed to attach at 37 8C for 18 –20 h. Subsequently, the cells were washed twice with 1 ml of PBS (composed of 137 mM NaCl, 2.7 mM KCl, 8 mM Na2HPO4 and 1.5 mM KH2PO4) and incubated in 1 ml of PBS, with or without test peptide (10, 100 or 1000 nM gigantoxin I or 0.01, 0.1, 1 or 10 nM human EGF), at 37 8C for 30 min. Morphological changes of the cells were examined by phase-contrast microscopy. For the assay of tyrosine phosphorylation of the EGFR, A431 cells (1 £ 106 cells) were grown in 1 ml of DMEM in 10 cm dishes at 37 8C for 18– 20 h. Following incubation with test peptide at 37 8C for 30 min, the cells were washed twice with cold PBS (supplemented with 0.9 mM CaCl2 and 0.5 mM MgCl2), scraped with a cell scraper into 1 ml of the same cold PBS and centrifuged at 1000 £ g and 4 8C for 4 min. The cell pellet, after being stored at 2 80 8C for 18 – 20 h, was homogenized in 1 ml of radioimmune precipitation (RIPA) buffer (25 mM HEPES buffer, pH 7.8, containing 1.5% Triton-X100, 0.1% SDS, 500 mM NaCl, 100 mM Na3VO4, 50 mM NaF, 5 mM EDTA, 1 mM phenylmethanesulfonyl fluoride, 0.01% leupeptin and 1% sodium deoxycholate) at 4 8C for 30 min and centrifuged at 15000 £ g and 4 8C for 15 min. The supernatant was mixed with 25 ml of 50% Protein G Sepharose 4B (Amersham Pharmacia Biotech) suspension in RIPA buffer and kept with stirring at 4 8C for 1 h. After centrifugation at 2000 £ g and 4 8C for 4 min, the supernatant was added with 2 mg of mouse anti-EGFR antibody (Santa Cruze Biotechnology, Santa Cruze, USA) and 25 ml of 50% Protein G Sepharose 4B suspension in RIPA buffer. The mixture was stirred at 4 8C for 3 h and centrifuged at 2000 £ g and 4 8C for 4 min. After being washed four times with 1 ml of RIPA buffer, the precipitate was suspended in 25 ml of electrophoresis loading buffer (0.52 M Tris – HCl buffer, pH 6.8, containing 2.15% SDS, 5% 2-mercaptoethanol, 7% glycerol and 0.0025% bromophenol blue; Toyobo, Tokyo, Japan), heated at 95 8C for 3 min and centrifuged briefly. A 20 ml portion of the supernatant was subjected to SDS – PAGE on a 5% polyacrylamide gel, which was performed by the method of Laemmli (1970). Proteins separated by SDS – PAGE were transferred onto a nitrocellulose membrane at 100 V for 1 h.
K. Shiomi et al. / Toxicon 41 (2003) 229–236
The membrane was blocked with 10% skimmed milk in TBS-Tween (10 mM Tris – HCl buffer, pH 7.5, containing 0.1% Tween 20) at room temperature for 1 h and washed with TBS-Tween. The transfer was probed with mouse antiphosphotyrosine antibody 4G10 (1:5000 dilution; Seikagaku Corp., Tokyo, Japan) at room temperature for 2 h, followed by anti-mouse IgG antibody RPN 1001 conjugated with horseradish peroxidase (1:1000 dilution; Amersham Pharmacia Biotech) at room temperature for 1 h. After rinsing with TBS, labeled proteins were visualized with an ECL detection system (Amersham Pharmacia Biotech). 2.5. Peptide determination Peptides were determined by the method of Lowry et al. (1951) using bovine serum albumin as a standard. 2.6. Reduction and S-pyridylethylation
231
3. Results 3.1. Isolation of gigantoxins In gel filtration on Sephadex G-50, toxins with crab lethality appeared between fractions 30 and 42 (Fig. 1(A)). When the toxic fraction obtained by gel filtration was subjected to reverse-phase HPLC on TSKgel ODS-120T, three toxins (named gigantoxins I, II and III in the order of elution) were eluted at retention times of 31, 39 and 50 min, respectively (Fig. 1(B)). In a typical run, 5 g of the starting sample yielded as much as 5960 mg of gigantoxin I, 150 mg of gigantoxin II and 250 mg of gigantoxin III. 3.2. Amino acid sequences of gigantoxins The entire amino acid sequences of gigantoxins I (48 residues), II (44 residues) and III (48 residues) were determined based on the strategies shown in Fig. 2.
Each purified gigantoxin (200 – 300 mg) was dissolved in 300 ml of 0.5 M Tris – HCL buffer (pH 8.5) containing 7 M guanidine hydrochloride and 10 mM EDTA and reduced with 1 mg of dithiothreitol at room temperature for 2 h. The reaction mixture was then added with 2 ml of 4-vinylpyridine and kept in the dark at room temperature for 1.5 h. The pyridylethylated (PE)-toxin was purified by reverse-phase HPLC on TSKgel ODS-120T with a linear gradient of acetonitrile (0 –70% in 40 min) in 0.1% TFA. 2.7. Enzyme digestion PE-gigantoxins (100 mg each) were individually digested with 40 mg of carboxypeptidase W (EC 3.4.16.1; Seikagaku Corp.) in 120 ml of 0.05 M acetate buffer (pH 4.0) at 30 8C. Aliquots of the reaction mixture were withdrawn at suitable intervals and subjected to an MLC-703 amino acid analyzer (Atto, Tokyo, Japan). PE-gigantoxin II (70 mg) was also digested with 5 mg of chymotrypsin (EC 3.4.21.1; Roche, Mannheim, Germany) in 500 ml of 0.1 M Tris – HCl buffer (pH 7.8) containing 0.01 M CaCl2 at 25 8C for 18 h and PE-gigantoxin III (50 mg) with 10 mg of V8 protease (EC 3.4.21.19; ICN Biomedicals, Irvine, USA) in 250 ml of 0.1 M ammonium hydrogencarbonate containing 2 mM EDTA at 37 8C for 18 h. Peptide fragments produced were isolated for sequencing by reverse-phase HPLC on TSKgel ODS-120T with a linear gradient of acetonitrile (0– 50% in 80 min) in 0.1% TFA. 2.8. Sequence analysis Amino acid sequence analyses were performed with an automatic gas-phase protein sequencer (LF-3400D TriCart with high sensitivity chemistry; Beckman Coulter, Fullerton, USA).
Fig. 1. Isolation of gigantoxins. (A) Gel filtration on Sephadex G50. Toxic fractions are indicated by a bar. (B) Reverse-phase HPLC on TSKgel ODS-120T. Gigantoxins I– III were eluted in labeled peaks.
232
K. Shiomi et al. / Toxicon 41 (2003) 229–236
Fig. 2. Strategies for the determination of entire amino acid sequences of gigantoxins I– III. Labels above the straight lines with arrows at both ends: Native, N-terminal sequences determined by sequencing of the native peptides; CpW, C-terminal sequences determined by the carboxypeptidase W digestion method; Chy, sequences of peptide fragments from the chymotrypsin digest of PE-gigantoxin II; V8, sequence of a peptide fragment from the V8 protease digest of PE-gigantoxin III.
Sequencing of the native peptides identified N-terminal 46, 35 and 45 residues for gigantoxins I, II and III, respectively. The C-terminal sequences were deduced to be 44EQMSV48 for gigantoxin I, 43KQ44 for gigantoxin II and 46RPR48 for
gigantoxin III by the estimation of amino acids released upon carboxypeptidase W digestion of the PE-derivatives. In the case of gigantoxin I, the N- and C-terminal sequences partly overlapped with each other. The remaining sequence
Fig. 3. Alignment of the amino acid sequence of gigantoxin I with those of mammalian EGFs (A), gigantoxin II with type 1 sea anemone sodium channel toxins (B) and gigantoxin III with type 2 sea anemone sodium channel toxins (C). (A) The residues identical with gigantoxin I are boxed. (B) AFT-I is from Anthopleura fuscoviridis (Sunahara et al., 1987); AP-A from A. xanthogrammica (Tanaka et al., 1977); ATX I from A. sulcata (Wunderer and Eulitz, 1978); Cp I from C. passiflora (Shiomi et al., 1995); and Rc I from Radianthus (Heteractis ) crispus (Shiomi et al., 1996). The residues identical with gigantoxin II are boxed. Asterisks under the sequences of Rc I represent the residues typical of type 1 toxins. Hydroxyproline at position 3 of Cp I and Rc I is indicated by ‘O’. (C) Rp II from R. paumotensis (Wemmer et al., 1986); RTX I (Zykova and Kozlovskaya, 1989), II (Zykova et al., 1988a) and V (Zykova et al., 1988b) from R. macrodactylus; and Sh I from S. helianthus (Kem et al., 1989). Asterisks under the sequences of Sh I represent the residues typical of type 2 toxins.
K. Shiomi et al. / Toxicon 41 (2003) 229–236
of gigantoxin II was compensated by two peptide fragments (corresponding to the segments 32 –41 and 42 – 44) isolated from the chymotrypsin digest of the PE-derivative. Although not confirmed, these two fragments were judged to be linked together, since, as described below, gigantoxin II is a member of the type 1 sea anemone sodium channel toxins with two successive cysteine residues in the Cterminal region. As for gigantoxin III, one of the peptide fragments isolated from the V8 protease digest of the PEderivative was unequivocally the C-terminal peptide (corresponding to the segment 29 –48) overlapping with the N-terminal sequence. Very surprisingly, a search by protein sequence databases revealed that gigantoxin I is very similar to mammalian EGFs. For comparison, the amino acid sequence of gigantoxin I is aligned with those of EGFs from human (Gregory and Preston, 1977), pig (Pascall et al., 1991) and mouse (Savage et al., 1972) in Fig. 3(A). As much as 31 – 33% homologies are recognized between gigantoxin I and EGFs; the location of six cysteine residues in gigantoxin I is completely identical with that in EGFs. On the other hand, gigantoxins II and III are obviously analogous to type 1 and 2 toxins, respectively, when aligned with the following known sea anemone sodium channel toxins (Fig. 3(B) and (C)): type 1 toxins, AFT-I from Anthopleura fuscoviridis (Sunahara et al., 1987), AP-A from Anthopleura xanthogrammica (Tanaka et al., 1977), ATX I from Anemonia sulcata (Wunderer and Eulitz, 1978), Cp I from Condylactis passiflora (Shiomi et al., 1995) and Rc I from Radianthus (Heteractis ) crispus (Shiomi et al., 1996); type 2 toxins, Rp II from Radianthus paumotensis (Wemmer et al., 1986), RTX I (Zykova and Kozlovskaya, 1989), II (Zykova et al., 1988a) and V (Zykova et al., 1988b) from Radianthus macrodactylus and Sh I from Stichodactyla helianthus (Kem et al., 1989). The residues typical of type 1 and two toxins are well conserved in gigantoxins II and III, respectively. It is, however, noticeable that gigantoxin II has a rather shorter chain than the known type 1 toxins and that gigantoxin III has an unusual C-terminal sequence (46RPR48) as compared to the known type 2 toxins.
233
human EGF, any ill signs including paralysis were not observed in both mice and crabs even at 1000 mg/kg. EGF, in Ca2þ free media, is known to change the polygonal pavement-like shape (Fig. 4(A)) of A431 cells to the round shape usually within 10 min (Chinkers et al., 1981). In this study, therefore, rounding of the cells was judged 30 min after treatment with human EGF or gigantoxin I. As a result, human EGF caused rounding of almost all A431 cells at 1 and 10 nM (data not shown), while gigantoxin I showed the same effect at a 100– 1000 times higher concentration (1000 nM) (Fig. 4(B)). No morphological changes were observed at less than 0.1 nM human EGF and less than 100 nM gigantoxin I. Furthermore, gigantoxin I was found to induce tyrosine phosphorylation of the EGFR in A431 cells. When the EGFR was purified from the cells incubated with the toxin by use of the anti-EGFR antibody and subjected to Western blotting, the 170 kDa protein corresponding to the EGFR reacted with the anti-phosphotyrosine antibody (Fig. 5). The intensity of the visualized band at 500 or 1000 nM gigantoxin I was almost the same as that at 1 nM human EGF, indicating that gigantoxin I is 500– 1000 times less in the potency of inducing autophosphorylation of the EGFR than human EGF.
3.3. Biological activities of gigantoxins None of the three gigantoxins caused any symptoms in mice at a dose of 1000 mg/kg. As for crab toxicity, both gigantoxins II and III were potently lethal with LD50 of 70 and 120 mg/kg, respectively. In contrast, gigantoxin I induced tonic paralysis in crabs with an ED50 of 215 mg/kg, while it killed only one of the five crabs challenged with a dose of 1000 mg/kg, indicating that its LD50 is over 1000 mg/kg. The crabs injected with effective doses near ED50 exhibited paralysis after 10 – 20 min, which made them impossible to walk for 5 – 10 min. At 1000 mg/kg, paralysis appeared immediately after injection and lasted for a longer time (30 –60 min). In the case of
Fig. 4. Morphological change of A431 cells by gigantoxin I. The cells were incubated at 37 8C for 30 min in PBS without (A) or with gigantoxin I (1000 nM) (B).
234
K. Shiomi et al. / Toxicon 41 (2003) 229–236
Fig. 5. Tyrosine phosphorylation of EGFR in A431 cells by human EGF and gigantoxin I. After the cells were incubated with human EGF or gigantoxin I at 37 8C for 30 min, the EGFR was purified from the cells and subjected to Western blot analysis. Lanes: 1, 10 nM human EGF; 2, 1 nM human EGF; 3, 1000 nM gigantoxin I; 4, 500 nM gigantoxin I; 5, 100 nM gigantoxin I; 6, control.
4. Discussion In this study, three peptide toxins (gigantoxins I– III) with crab toxicity were isolated from the sea anemone S. gigantea and their complete amino acid sequences were determined. One peptide toxin with crab lethality has previously been isolated from the same sea anemone (formerly called Stoichactis giganteus ) but its amino acid sequence has not been elucidated (Schweitz et al., 1981). Judging from the reported amino acid composition, this peptide toxin seems to be the same as gigantoxin III. Of the three peptide toxins from S. gigantea, gigantoxin I is of particular interest in that it has high homologies (31 – 33%) with mammalian EGFs with the same location of six cysteine residues. In accordance with the sequence similarity, gigantoxin I induces the rounding of A431 cells and tyrosine phosphorylation of the EGFR in the cells, although much less potently than human EGF. However, gigantoxin I is distinguishable from EGFs in crab toxicity; gigantoxin I is weakly lethal with LD50 . 1000 mg/kg and causes paralysis with an ED50 of 215 mg/kg, while human EGF is non-toxic even at 1000 mg/kg. As far as we know, gigantoxin I is the first example of EGF-like toxins of natural origin. EGF-like toxins such as gigantoxin I may be contained in sea anemones other than S. gigantea. In future survey of EGF-like toxins in sea anemones, however, it should be kept in mind that they may be paralytic but not potently lethal to crabs, as observed with gigantoxin I. According to Carpenter and Cohen (1990), EGFs and EGF-like molecules such as transforming growth factor-a have the same residues at 11 positions, i.e. Cys-6 (numbering is based on the sequence of gigantoxin I shown in Fig. 2), Cys-15, Gly-19, Cys-21, Cys-32, Cys-34, Tyr-38, Gly-40, Arg-42, Cys-43 and Leu-48. All of these residues, except for Leu-48, are also well conserved in gigantoxin I. It should be noted that Leu-48 has previously
been identified as one of the essential residues for the biological activity of EGFs (Engler et al., 1988; Ray et al., 1988). Thus, the absence of Leu-48 seems to explain the reason why gigantoxin I is much less potent in EGF activity (rounding of A431 cells and tyrosine phosphorylation of the EGFR) than human EGF. It is also interesting to note that gigantoxin I has marked alterations including the absence of Leu-48 in the C-terminal portion as compared to EGFs. This may be the molecular basis for the crab toxicity of gigantoxin I. Sea anemone toxins are generally considered to be contained in specialized stinging cells called nematocysts. This is probably the case with gigantoxins, although not confirmed. In relation to this, it is worth mentioning that the content of gigantoxin I in S. gigantea is assumed to be more than 0.1% on the basis of yield (5960 mg from 5 g of wet sample), being much higher than those of gigantoxins II and III. Although gigantoxin I is substantially non-lethal to crabs, its high content and potent paralytic activity against crabs suggest that it functions in S. gigantea as a more powerful weapon to paralyze prey animals than the other two peptide toxins. The finding of an EGF-like toxin (gigantoxin I) in S. gigantea is also of interest from the viewpoint of molecular evolution of EGFs. So far, there have been numerous studies on EGFs from wide species of animals, especially vertebrates. Nevertheless, no information is available concerning EGFs in animals, such as sponges, corals and sea anemones, which are the nearest to the phylogenetic root of the animal kingdom. The present study clearly demonstrates that gigantoxin I belongs to the EGF family with respect to structure and activity. However, the significantly high content of gigantoxin I in S. gigantea is extremely disproportionate to the known regulatory molecules such as mammalian EGFs. This suggests that gigantoxin I seems not to be important as a growth factor for the sea anemone. It is thus tentatively assumed that the ancestors of EGFs originally had functioned as toxins as in the case of gigantoxin I and that they had lost toxic properties during the evolution process in the animal kingdom. In order to ascertain this fascinating assumption, it is essential to elucidate the gene structure of gigantoxin I and compare it with those of the known EGFs. The gene expression analysis probed with the gigantoxin I cDNA sequence is also necessary to know if gigantoxin I is the sole EGF-like molecule in the sea anemone or another EGF is present. As for gigantoxins II and III, they are analogous to the known sea anemone sodium channel toxins. According to the classification of sea anemone sodium channel toxins by Norton (1991), gigantoxins II and III are identified as type 1 and 2 toxins, respectively. It has previously been proposed that sea anemones belonging to the genera Anthopleura and Anemonia contain type 1 toxins, while those to the genera Radianthus (or Heteractis ) and Stichodactyla contain type 2 toxins (Norton, 1991). Radianthus crispus not having a type 2 toxin but a type 1 toxin has been the only exception
K. Shiomi et al. / Toxicon 41 (2003) 229–236
(Shiomi et al., 1996). Therefore, our results are valuable in showing first the coexistence of both type 1 and 2 toxins in one species of sea anemone. The relationship between the taxonomical position of sea anemones and the type of their sodium channel toxins will require future detailed study. Finally, it should be again emphasized that gigantoxin I is the first example of EGF-like molecules with both toxic and EGF activities. Its mode of action on crabs, structure – activity relationships and gene structure remain to be elucidated. Also, the distribution of similar EGF-like toxins in sea anemones should be surveyed.
Acknowledgements The authors thank Miss S. Kawahata for technical assistance. This study was partly supported by a Grant-inAid for Scientific Research from the Ministry of Education, Science, Sports and Culture of Japan.
References Carpenter, G., Cohen, S., 1990. Epidermal growth factor. J. Biol. Chem. 265, 7709–7712. Castan˜eda, O., Sotolongo, V., Amor, A.M., Sto¨klin, R., Anderson, ˚ ., Wernstedt, C., Karlsson, E., A.J., Harvey, A.L., Engstro¨m, A 1995. Characterization of a potassium channel toxin from the Caribbean sea anemone Stichodactyla helianthus. Toxicon 33, 603–613. Chinkers, M., Mckanna, J.A., Cohen, S., 1981. Rapid rounding of epidermoid carcinoma cells A-431 induced by epidermal growth factor. J. Cell Biol. 88, 422–429. Cotton, J., Crest, M., Bouet, F., Alessandri, N., Gola, M., Forest, E., Karlsson, E., Castan˜eda, O., Harvey, A.L., Vita, C., Me´nez, A., 1997. A potassium-channel toxin from the sea anemone Bunodosoma granulifera, an inhibitor for Kv1 channels. Revision of the amino acid sequence, disulfide-bridge assignment, chemical synthesis, and biological activity. Eur. J. Biochem. 244, 192–202. Diochot, S., Schweitz, H., Be´ress, L., Lazdunski, M., 1998. Sea anemone peptides with a specific blocking activity against the fast inactivating potassium channel Kv3.4. J. Biol. Chem. 273, 6744–6749. Engler, D.A., Matsunami, R.K., Campion, S.R., Stringer, C.D., Stevens, A., Niyogi, S.K., 1988. Cloning of authentic human epidermal growth factor as a bacterial secretary protein and its initial structure–function analysis by site-directed mutagenesis. J. Biol. Chem. 263, 12384–12390. Gendeh, G.S., Young, L.C., de Medeiros, L.C., Jeyaseelan, K., Harvey, A.L., Chung, M.C.M., 1997. A new potassium channel toxin from the sea anemone Heteractis magnifica: isolation, cDNA cloning, and functional expression. Biochemistry 36, 11461– 11471. Gregory, H., Preston, B.M., 1977. The primary structure of human urogastrone. Int. J. Pept. Protein Res. 9, 107–118. Ishida, M., Yokoyama, A., Shimakura, K., Nagashima, Y., Shiomi, K., 1997. Halcurin, a polypeptide toxin in the sea anemone
235
Halcurias sp with a structural resemblance to type 1 and 2 toxin. Toxicon 35, 537–544. Kem, W.R., 1988. Sea anemone toxins: structure and action. In: Hessinger, D.A., Lenhoff, H.M. (Eds.), The Biology of Nematocysts, Academic Press, New York, pp. 375 –405. Kem, W.R., Parten, B., Pennington, M.W., Price, D.A., Dunn, B.M., 1989. Isolation, characterization, and amino acid sequence of a polypeptide neurotoxin occurring in the sea anemone Stichodactyla helianthus. Biochemistry 28, 3483–3489. Kem, W.R., Pennington, M.W., Dunn, B.M., 1990. Sea anemone polypeptide toxins affecting sodium channels. In: Hall, S., Stricharz, G. (Eds.), Marine Toxins. Origin, Structure and Molecular Pharmacology, ACS, Washington, DC, pp. 279 –289. Kem, W.R., Pennington, M.W., Krafte, D.S., Hill, R.J., 1996. Sea anemone toxins affecting sodium channels: are the similarities than the differences? . In: Lazarovici, P., Spira, M.E., Zlotkin, E. (Eds.), Biochemical Aspects of Marine Pharmacology, Alakasn, Fort Collins, pp. 98–120. Laemmli, U.K., 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680 –685. Lin, X.Y., Ishida, M., Nagashima, Y., Shiomi, K., 1996. A polypeptide toxin in the sea anemone Actinia equina homologous with other sea anemone sodium channel toxins: isolation and amino acid sequence. Toxicon 34, 57–65. Litchfield, J.T. Jr, Wilcoxon, F., 1949. A simplified method of evaluating dose-effect experiments. J. Pharmacol. Exp. Ther. 96, 99–113. Lowry, O.H., Rosebrough, A.L., Farr, A.L., Randall, R.J., 1951. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193, 265–275. Minagawa, S., Ishida, M., Nagashima, Y., Shiomi, K., 1998. Primary structure of a potassium channel toxin from the sea anemone Actinia equina. FEBS Lett. 427, 149– 151. Norton, R.S., 1991. Structure and structure–function relationships of sea anemone proteins that interact with the sodium channel. Toxicon 29, 1051–1084. Pascall, J.C., Jones, D.S.C., Doel, S.M., Clements, J.M., Hunter, M., Fallon, T., Edwards, M., Brown, K.D., 1991. Cloning and characterization of a gene encoding pig epidermal growth factor. J. Mol. Endocrinol. 6, 63–70. Ray, P., Mey, F.J., Montelione, G.T., Liu, J.-F., Narang, S.A., Scherage, H.A., Wu, R., 1988. Structure–function studies of murine epidermal growth factor: expression and site-directed mutagenesis of epidermal growth factor gene. Biochemistry 27, 7289–7295. Savage, C.R. Jr, Inagami, T., Cohen, S., 1972. The primary structure of epidermal growth factor. J. Biol. Chem. 247, 7612–7621. Schweitz, H., Vincent, J.P., Barhanin, J., Frelin, C., Linden, G., Hugues, M., Lazdunski, M., 1981. Purification and pharmacological properties of eight sea anemone toxins from Anemonia sulcata, Anthopleura xanthogrammica, Stoichactis giganteus, and Actinodendron plumosum. Biochemistry 20, 5245–5252. Schweitz, H., Bruhn, T., Guillemare, E., Moinier, D., Lancelin, J.M., Be´ress, L., Lazdunski, M., 1995. Kalicludines and kaliseptine. Two different classes of sea anemone toxins for voltage-sensitive Kþ channels. J. Biol. Chem. 270, 25121–25126.
236
K. Shiomi et al. / Toxicon 41 (2003) 229–236
Shiomi, K., Lin, X.Y., Nagashima, Y., Ishida, M., 1995. Isolation and amino acid sequence of polypeptide toxins in the sea anemone Condylactis passiflora. Fish. Sci. 61, 1016–1021. Shiomi, K., Lin, X.Y., Nagashima, Y., Ishida, M., 1996. Isolation and amino acid sequence of a polypeptide toxin from the sea anemone Radianthus crispus. Fish. Sci. 62, 629– 633. Shiomi, K., Qian, W.H., Lin, X.Y., Shimakura, K., Nagashima, Y., Ishida, M., 1997. Novel polypeptide toxins with crab toxicity from the sea anemone Anemonia erythraea. Biochim. Biophys. Acta 1335, 191 –198. Sunahara, S., Muramoto, K., Tenma, K., Kamiya, H., 1987. Amino acid sequence of two sea anemone toxins from Anthopleura fuscoviridis. Toxicon 25, 211– 219. Tanaka, M., Haniu, M., Yasunobu, K.T., Norton, T.R., 1977. Amino acid sequence of the Anthopleura xanthogrammica heart stimulant anthopleurin-A. Biochemistry 16, 204–208.
Wemmer, D.E., Kumar, N.V., Metrione, R.M., Lazdunski, M., Drobny, G., Kallenbach, N.R., 1986. NMR analysis and sequence of toxin II from the sea anemone Radianthus paumotensis. Biochemistry 25, 6842– 6849. Wunderer, G., Eulitz, M., 1978. Amino acid sequence of toxin 1 from Anemonia sulcata. Eur. J. Biochem. 89, 11–17. Zykova, T.A., Kozlovskaya, E.P., 1989. Amino acid sequence of neurotoxin I from the sea anemone Radianthus macrodactylus. Bioorg. Khim. 15, 1301–1306. Zykova, T.A., Kozlovskaya, E.P., Elyakov, G.B., 1988a. Amino acid sequence of neurotoxin II from the sea anemone Radianthus macrodactylus. Bioorg. Khim. 14, 878 –882. Zykova, T.A., Kozlovskaya, E.P., Elyakov, G.B., 1988b. Amino acid sequence of neurotoxins IV and V from the sea anemone Radianthus macrodactylus. Bioorg. Khim. 14, 1489–1494.